April 2015 - Devel - libvirt List Archives

[libvirt] [PATCH] parallels: use SDK disks system names for paths
by Nikolay Shirokovskiy 20 Apr '15

20 Apr '15

For block devices SDK friendly name do not refer to block device path but rather some description while system name refer to device path. For ploop devices both names refer to image path. Thus system name is better choice. Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy(a)parallels.com> --- src/parallels/parallels_sdk.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/parallels/parallels_sdk.c b/src/parallels/parallels_sdk.c index 72c1773..e1ba7cf 100644 --- a/src/parallels/parallels_sdk.c +++ b/src/parallels/parallels_sdk.c @@ -485,13 +485,13 @@ prlsdkGetDiskInfo(PRL_HANDLE prldisk, disk->device = VIR_DOMAIN_DISK_DEVICE_DISK; } - pret = PrlVmDev_GetFriendlyName(prldisk, NULL, &buflen); + pret = PrlVmDev_GetSysName(prldisk, NULL, &buflen); prlsdkCheckRetGoto(pret, cleanup); if (VIR_ALLOC_N(buf, buflen) < 0) goto cleanup; - pret = PrlVmDev_GetFriendlyName(prldisk, buf, &buflen); + pret = PrlVmDev_GetSysName(prldisk, buf, &buflen); prlsdkCheckRetGoto(pret, cleanup); if (virDomainDiskSetSource(disk, buf) < 0) -- 1.7.1

1 0

[libvirt] [PATCH v4 0/9] qemu: Add quorum support to libvirt
by Matthias Gatto 20 Apr '15

20 Apr '15

The purpose of these patches is to introduce quorum for libvirt I've try to follow this proposal: http://www.redhat.com/archives/libvir-list/2014-May/msg00533.html This feature ask for 6 task: 1) Allow a _virStorageSource to contain more than one backing store. Because all the actual libvirt code use the backingStore field as a pointer and we needs want to change that, I've decide to encapsulate the backingStore field to simplifie the array manipulation. 2) Add the missing field a quorum need in _virStorageSource and the VIR_STORAGE_TYPE_QUORUM and VIR_STORAGE_FILE_QUORUM in their respectives enums. 3) Parse and format the xml Because a quorum allows to have more than one backing store at the same level we need to change virDomainDiskDefFormat and virDomainDiskDefParseXML to call virDomainDiskBackingStoreFormat and virDomainDiskBackingStoreParse in a loop. virDomainDiskBackingStoreFormat and virDomainDiskBackingStoreParse can call themself recursively in a loop because a quorum can contain another quorum 4) Add nodename We need to add nodename support in _virStorageSource because qemu use them for their child. 5) Build qemu string As for the xml, we have to call the function which create quorum recursively. But this task have the problem explained here: http://www.redhat.com/archives/libvir-list/2014-October/msg00529.html The _virStorageSource missing some informations that can be passed to a child, and therefore this version of quorum is incomplet. 6) Allow to hotplug/change a disk in a quorum This part is not present in these patches because for this task we have to use blockdev-add, and currently libvirt use device_add for hotpluging that doesn't allow to hotplug quorum childs. There is 3 way to handle this problem: 1) create a virDomainBlockDevAdd function in libvirt witch call blockdev-add. 2) use blockdev-add instead of device_add in qemuMonitorJSONAddDevice 3) write a hack which uses blockdev-add when only attaching quorum (but i'm pretty sure this solution is not the good one) V2: -Rebase on master -Add Documentation V3: -Transforme the backingStore field in virStorageSource into an array of pointer instead of a pointer -Modify virStorageSourceSetBackingStore to allow it to expand the backingStore size. V4: -Rebase on master Matthias Gatto (9): virstoragefile: Add virStorageSourceGetBackingStore virstoragefile: Always use virStorageSourceGetBackingStore to get backing store virstoragefile: Add virStorageSourceSetBackingStore virstoragefile: Always use virStorageSourceSetBackingStore to set backing store virstoragefile: change backingStore to backingStores. virstoragefile: Add quorum in virstoragefile domain_conf: Read and Write quorum config qemu: Add quorum support in qemuBuildDriveDevStr virstoragefile: Add node-name docs/formatdomain.html.in | 27 ++++- docs/schemas/domaincommon.rng | 95 +++++++++++------ docs/schemas/storagecommon.rng | 1 + docs/schemas/storagevol.rng | 1 + src/conf/domain_conf.c | 195 ++++++++++++++++++++++++++-------- src/conf/storage_conf.c | 22 ++-- src/libvirt_private.syms | 2 + src/qemu/qemu_cgroup.c | 4 +- src/qemu/qemu_command.c | 114 ++++++++++++++++++++ src/qemu/qemu_domain.c | 2 +- src/qemu/qemu_driver.c | 30 +++--- src/qemu/qemu_migration.c | 1 + src/security/security_dac.c | 2 +- src/security/security_selinux.c | 4 +- src/security/virt-aa-helper.c | 2 +- src/storage/storage_backend.c | 35 +++--- src/storage/storage_backend_fs.c | 37 ++++--- src/storage/storage_backend_gluster.c | 10 +- src/storage/storage_backend_logical.c | 15 ++- src/storage/storage_driver.c | 2 +- src/util/virstoragefile.c | 116 +++++++++++++++++--- src/util/virstoragefile.h | 13 ++- tests/virstoragetest.c | 18 ++-- 23 files changed, 573 insertions(+), 175 deletions(-) -- 1.8.3.1

3 16

[libvirt] [PATCH] parallels: use SDK disks system names for paths
by Nikolay Shirokovskiy 20 Apr '15

20 Apr '15

For block devices SDK friendly name do not refer to block device path but rather some description while system name refer to device path. For ploop devices both names refer to image path. Thus system name is better choice. Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy(a)parallels.com> --- src/parallels/parallels_sdk.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/parallels/parallels_sdk.c b/src/parallels/parallels_sdk.c index 80e9d00..4911ab7 100644 --- a/src/parallels/parallels_sdk.c +++ b/src/parallels/parallels_sdk.c @@ -491,7 +491,7 @@ prlsdkGetDiskInfo(PRL_HANDLE prldisk, if (VIR_ALLOC_N(buf, buflen) < 0) goto cleanup; - pret = PrlVmDev_GetFriendlyName(prldisk, buf, &buflen); + pret = PrlVmDev_GetSysName(prldisk, buf, &buflen); prlsdkCheckRetGoto(pret, cleanup); if (virDomainDiskSetSource(disk, buf) < 0) -- 1.7.1

2 1

[libvirt] [PATCH] rng: Forbid to validate mismatched <disk> 'device' and 'type' attributes
by Erik Skultety 20 Apr '15

20 Apr '15

According to docs, using 'lun' as a value for device attribute is only valid with disk types 'block' and 'network'. However current RNG schema also allows a combination type='file' device='lun' which results in a successfull xml validation, but fails at qemuBuildCommandLine. Besides fixing the RNG schema, this patch also adds a qemuxml2argvtest for this case. https://bugzilla.redhat.com/show_bug.cgi?id=1210669 --- docs/schemas/domaincommon.rng | 42 ++++++++++++++-------- .../qemuxml2argv-disk-device-lun-type-invalid.xml | 28 +++++++++++++++ tests/qemuxml2argvtest.c | 2 ++ 3 files changed, 58 insertions(+), 14 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-device-lun-type-invalid.xml diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng index 03fd541..ee32b73 100644 --- a/docs/schemas/domaincommon.rng +++ b/docs/schemas/domaincommon.rng @@ -1316,37 +1316,43 @@ </choice> </attribute> </optional> + <interleave> + <ref name="diskSource"/> + <ref name="diskSpecsExtra"/> + </interleave> </group> <group> <attribute name="device"> - <choice> - <value>lun</value> - </choice> + <value>lun</value> </attribute> <optional> <ref name="rawIO"/> </optional> <optional> - <attribute name="sgio"> - <choice> - <value>filtered</value> - <value>unfiltered</value> - </choice> - </attribute> + <ref name="sgIO"/> </optional> + <interleave> + <choice> + <ref name="diskSourceNetwork"/> + <ref name="diskSourceBlock"/> + </choice> + <ref name="diskSpecsExtra"/> + </interleave> </group> </choice> <optional> <ref name="snapshot"/> </optional> - <interleave> - <ref name="diskSource"/> - <ref name="storageSourceExtra"/> - <ref name="diskBackingChain"/> - </interleave> </element> </define> + <define name="diskSpecsExtra"> + <interleave> + <ref name="storageSourceExtra"/> + <ref name="diskBackingChain"/> + </interleave> + </define> + <define name="diskBackingChain"> <choice> <ref name="diskBackingStore"/> @@ -5315,4 +5321,12 @@ <ref name="virYesNo"/> </attribute> </define> + <define name="sgIO"> + <attribute name="sgio"> + <choice> + <value>filtered</value> + <value>unfiltered</value> + </choice> + </attribute> + </define> </grammar> diff --git a/tests/qemuxml2argvdata/qemuxml2argv-disk-device-lun-type-invalid.xml b/tests/qemuxml2argvdata/qemuxml2argv-disk-device-lun-type-invalid.xml new file mode 100644 index 0000000..ed3c89e --- /dev/null +++ b/tests/qemuxml2argvdata/qemuxml2argv-disk-device-lun-type-invalid.xml @@ -0,0 +1,28 @@ +<domain type='qemu'> + <name>QEMUGuest1</name> + <uuid>c7a5fdbd-edaf-9455-926a-d65c16db1809</uuid> + <memory unit='KiB'>219100</memory> + <currentMemory unit='KiB'>219100</currentMemory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='i686' machine='pc'>hvm</type> + <boot dev='hd'/> + </os> + <clock offset='utc'/> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/bin/qemu</emulator> + <disk type='file' device='lun' rawio='yes'> + <driver name='qemu' type='qcow2'/> + <source file='/dev/HostVG/QEMUGuest1'/> + <target dev='vda' bus='virtio'/> + <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> + </disk> + <controller type='usb' index='0'/> + <controller type='ide' index='0'/> + <controller type='pci' index='0' model='pci-root'/> + <memballoon model='virtio'/> + </devices> +</domain> diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index 8d0a4aa..63bf6ff 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -817,6 +817,8 @@ mymain(void) QEMU_CAPS_DRIVE, QEMU_CAPS_DRIVE_FORMAT); DO_TEST("disk-drive-no-boot", QEMU_CAPS_DRIVE, QEMU_CAPS_DEVICE, QEMU_CAPS_BOOTINDEX); + DO_TEST_FAILURE("disk-device-lun-type-invalid", + QEMU_CAPS_DRIVE, QEMU_CAPS_DEVICE, QEMU_CAPS_VIRTIO_SCSI); DO_TEST("disk-usb", NONE); DO_TEST("disk-usb-device", QEMU_CAPS_DRIVE, QEMU_CAPS_DEVICE, QEMU_CAPS_DEVICE_USB_STORAGE, -- 1.9.3

2 3

[libvirt] [PATCH 0/2] Misc fixes
by Cédric Bosdonnat 20 Apr '15

20 Apr '15

Hi all, Here are 2 small fixes: the first one just gets the apparmor right for SLES (and openSUSE). The second patch fixes a problem with lxc-enter-namespace on pretty recent kernels. They surely have nothing to do with each other ;) Cédric Bosdonnat (2): Apparmor qemu abstraction fixes for SLES Open /proc/PID/ns/* read-only to avoid getting permission denied examples/apparmor/libvirt-qemu | 9 +++++++++ src/util/virprocess.c | 2 +- 2 files changed, 10 insertions(+), 1 deletion(-) -- 2.1.4

4 5

[libvirt] [PATCH] qemu: lifecycle: guest get rebooted after reboot and shutdown
by zhang bo 20 Apr '15

20 Apr '15

Steps: 1 virsh reboot guest1 --mode=acpi 2 virsh shutdown guest1 --mode=agent Expected result: As the SHUTDOWN job is after REBOOT, we expected the guest to be *shutoff*. Exacted result: After the 2 steps above, the guest got *rebooted*. The reason to this problem: 1 in qemuDomainReboot(mode acpi), it sets priv->fakeReboot to 1. 2 after shutdown/reboot, qemu monitor IO trigged qemuProcessHandleShutdown(), which finds that priv->fakeReboot is 1, and reboot the guest. Root Cause of the problem: After further look into the problem, We found that the design of acpi/agent shutdown/reboot seems a little chaotic. ----------------------------------- sheet1 who sets fakeReboot ----------------------------------- shutdown reboot acpi Y(0) Y(1) agent N N It's apparently, *acpi-mode* jobs set fakeReboot. ----------------------------------- sheet2 who needs to check fakeReboot(qemuProcessHandleShutdown()) ----------------------------------- shutdown reboot acpi Y Y agent *Y* *N* Things become a little odd here. only agent-mode reboot doesn't check fakeReboot. We can tell from the above 2 sheets, that they're not consistent. *Agent-mode shutdown checks fakeReboot(trigger by SHUTDOWN monitorIO event), while it didn't set it before.* The chaos is not caused by libvirtd, it's a systematic problem, including guest os and qemu. (guest os writes ACPI, triggers qemu, qemu then triggers libvirtd) Solution: A simple solution is to make the 1st sheet consistent to the 2nd sheet, which is: ----------------------------------- sheet3 who should set fakeReboot: ----------------------------------- shutdown reboot acpi Y(0) Y(1) agent *Y(0)* N ----------------------------------- we let agent-mode shutdown set fakeReboot to 0. Signed-off-by: Zhang Bo <oscar.zhangbo(a)huawei.com> Signed-off-by: Wang Yufei <james.wangyufei(a)huawei.com> --- src/qemu/qemu_driver.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 7eb5a7d..c751dcf 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -1959,6 +1959,8 @@ static int qemuDomainShutdownFlags(virDomainPtr dom, unsigned int flags) goto endjob; } + qemuDomainSetFakeReboot(driver, vm, isReboot); + if (useAgent) { qemuDomainObjEnterAgent(vm); ret = qemuAgentShutdown(priv->agent, agentFlag); @@ -1970,7 +1972,6 @@ static int qemuDomainShutdownFlags(virDomainPtr dom, unsigned int flags) */ if (!useAgent || (ret < 0 && (acpiRequested || !flags))) { - qemuDomainSetFakeReboot(driver, vm, isReboot); /* Even if agent failed, we have to check if guest went away * by itself while our locks were down. */ -- 1.7.12.4

2 2

[libvirt] Selective block device migration implementation
by Pavel Boldin 18 Apr '15

18 Apr '15

Dear Libvirt Developers, I'm working to implement feature request [1]. The feature request proposes to enhance `libvirt' code so the API caller can specify which block devices are to be migrated using e.g. parameters in the `virDomainMigrateToURI3' call. There is the following issues: 1. It is obvious to identify devices to be migrated using the `device_name'. However, these need to be serialized in either 1) a comma-separated string or in 2) a set of values named like `blockdevice%d' with `blockdevice' value holding amount of block devices to migrate. What is the desired approach here? Can block device name contain commas or should I neglect this possibility? 2. `libvirt' behavior in block devices migration was copied from the `QEMU' implementation that ignores read-only devices as well. So, `libvirt' code will need to pass this list to `QEMU' `migrate' QMP command and this argument should be implemented in the `QEMU'. Is there any implementation advices from your side? Pavel [1] https://bugzilla.redhat.com/show_bug.cgi?id=1203032

2 5

[libvirt] converting gatewaydev= from domU.cfg to libvirt.xml fails
by Olaf Hering 18 Apr '15

18 Apr '15

How should this be converted? /etc/init.d/boot.local tap=xentap tunctl -pt ${tap} ip addr add 1.1.1.1/29 dev ${tap} ip link set up dev ${tap} domU.cfg vif=[ 'mac=00:16:3e:13:01:00,ip=1.1.1.2,type=vif,gatewaydev=xentap,script=vif-route' ] The result from "convert-xml xen-xl domU,cfg" is: <interface type='ethernet'> <mac address='00:16:3e:13:01:00'/> <ip address='1.1.1.2' family='ipv4'/> <script path='vif-route'/> </interface> gatewaydev= is missing, the guest will not start. Olaf

2 2

[libvirt] [libvirt-php] memory overflow when using libvirt_list_networks
by Kash Pande 18 Apr '15

18 Apr '15

Finding lots of bad behaviour here: [Fri Apr 17 23:44:04.590626 2015] [:error] [pid 31714] [client 127.0.0.1:38103] PHP Warning: libvirt_list_networks(): this function is not supported by the connection driver: virConnectNumOfNetworks [Fri Apr 17 23:44:04.590655 2015] [:error] [pid 31714] [client 127.0.0.1:38103] PHP Fatal error: Out of memory (allocated 524288) (tried to allocate 18446744073709551608 bytes) There's no way to check if this exists... I noticed the following about versions: - this system works fine (Git): libvirt 1.2.9 Array ( [libvirt.release] => 9 [libvirt.minor] => 2 [libvirt.major] => 1 [connector.version] => 0.5.0 [connector.major] => 0 [connector.minor] => 5 [connector.release] => 0 ) - this system has problems (Git) libvirt 1.2.12 Array ( [libvirt.release] => 12 [libvirt.minor] => 2 [libvirt.major] => 1 [connector.version] => 0.5.0 [connector.major] => 0 [connector.minor] => 5 [connector.release] => 0 [type.release] => 12 [type.minor] => 2 [type.major] => 1 ) When I do a version check to only do network list for version under .12, I get this: [Fri Apr 17 23:44:43.608308 2015] [:error] [pid 31689] [client 127.0.0.1:38216] PHP Warning: libvirt_network_get(): this function is not supported by the connection driver: virNetworkLookupByName Which causes a cascade of other errors if you're not checking for problems each step along the way. Kash Pande Jentu Technologies, Inc http://jentu-networks.com

1 0

[libvirt] [PATCH v3 0/4] storage: handle scsi/iscsi error paths better
by John Ferlan 18 Apr '15

18 Apr '15

Rework of unpushed patches from series: http://www.redhat.com/archives/libvir-list/2015-April/msg00050.html Details: Patch 1 - Split out the non /dev/* checks into their own to start satisfying Jan's comment in his review of the former patch 3: http://www.redhat.com/archives/libvir-list/2015-April/msg00100.html Patch 2 - New - Flip the logic from assume success, change status to failure, and return to assume failure and only have success when everything completes successfully Patch 3 - Adjustments for previous patch 3 to make use of patch 1's split of the checking for using the /dev path. If we find that the /dev/* is not in use for the pool, then force an error earlier. This makes the check later on not need to compare the returned path to the generated 'devpath' to compensate for the logic in virStorageBackendStablePath which will strdup the incoming target path if something goes wrong in looking up the directory. Since both paths will use the new virStorageBackendPoolUseDevPath slightly differently, if devpath is the same as vol->target.path, then we know it's not because the pool target path was to blame. Patch 4 - Is the former patch 5, but now since patch 3 returns a failure in virStorageBackendSCSINewLun for that bad path, we can use it to be the error arbiter and bug fixer. Former patch 6 - no longer necessary, although it would circumvent any of the above patches being necessary to solve the bug... John Ferlan (4): storage: Split out the valid path check refactor virStorageBackendSCSINewLun scsi: Adjust return value for virStorageBackendSCSINewLun scsi: Adjust return values from processLU src/storage/storage_backend.c | 42 +++++++++---- src/storage/storage_backend.h | 1 + src/storage/storage_backend_scsi.c | 124 +++++++++++++++++++++---------------- 3 files changed, 100 insertions(+), 67 deletions(-) -- 2.1.0

2 14